Description of problem: When accessing a Volume larger than 4tb through dm-multipath I get [root@backup]/var/log> fsck /dev/san/estor02lun0p1 fsck 1.37 (21-Mar-2005) e2fsck 1.37 (21-Mar-2005) The filesystem size (according to the superblock) is 1202399223 blocks The physical size of the device is 128657399 blocks Either the superblock or the partition table is likely to be corrupt! Without multipath the device is accessible and fsck'ble. I suspect a 32bit-wrap because of [root@backup]~> bc bc 1.06 Copyright 1991-1994, 1997, 1998, 2000 Free Software Foundation, Inc. This is free software with ABSOLUTELY NO WARRANTY. For details type `warranty'. # 4k blocks to volume size 1202399223 * 4 4809596892 # wrap 4809596892 - (2^32) 514629596 # back to blocks 514629596 / 4 # qed... 128657399 My device map for this device is [root@backup]/~> multipath -l 3600d0230000000000083591ea0567c00 3600d0230000000000083591ea0567c00dm-5 IFT,A24F-R2224-1 [size=4.5T][features=0][hwhandler=0] \_ round-robin 0 [prio=0][active] \_ 3:0:1:0 sdp 8:240 [active][undef] (I don't have actually a multipath for this device, but I'd like to have...) Version-Release number of selected component (if applicable): Kernel 2.6.17 mutlipath-tools 0.4.7 How reproducible: - Create a SAN volume > 4tb (maybe 2tb) - Access it via (one of) its pysical devive => works - Access it through a setup with dm-multipath => volume size is mod 2 or 4tb (haven't tried a volume > 2tb and < 4tb yet)
We don't fix bugs that are not filed against a rhel4 kernel. If this is an upstream bug, please file it against fedora.